The ph2012 database is the last version of the meta-collection of puzzles with highest SER.
It has been defined here: http://forum.enjoysudoku.com/post296155.html#p296155
It has 3,103,972 puzzles, ranging from SER 10.2 to 11.9
Long ago, in support of my old T&E(2) conjecture, I had checked that all of these puzzles are at most in T&E(2), and therefore none is in T&E(3).
Considering the range of the SER, it should be expected that none is in T&E(1).
That's what I've just checked in collaboration with François Cordoliani (Defise). Except for the following two puzzles, which are in T&E(1):
- Code: Select all
.6.7.........8......5..9.1.8...6......9.25.3..7......4.....35.1....9..2...2...3..;11.00;1.20;1.20;FLK;final2 2018_07;2112870;21; #756628; real SER = 9.2
.6.7.....4...8......5....1.8...6......9..5.3..7.9....4.....35.1...49..2...2...3..;10.60;10.60;9.40;FLK;final2 2018_07;2112865;22; #1900698; real SER = 9.4
As you can see, they shouldn't be in the database at all: their SERs are grossly overestimated.